The major objective of the proposed research is to provide a more complete explanation of why Mexican Americans, who, based on most conventional indicators, represent a high-risk population, have such relatively favorable pregnancy outcomes. The research has the following specific aims: (i) to develop and test hypotheses regarding socioeconomic, demographic, sociocultural, biomedical, and physiological risk factors that may help to resolve the enigma of favorable pregnancy outcomes, especially birth weight, in the Mexican American population; (2) to explore these hypotheses in a comparative context that involves both non-Hispanic blacks and non-Hispanic whites; (3) to investigate the extent to which the birth outcome paradox "explains itself" in terms of bias introduced by ethnic selectivity of outcomes; (4) to examine, for the first time, whether a somewhat different set of risk factors is required to account for variation in the two components of low birth weight (preterm birth and intrauterine growth retardation) among Mexican Americans as appears to be the case among non-Hispanic blacks and whites. The rationale for the tremendous emphasis placed on birth weight by researchers, medical practitioners, and policy-makers resides initially in the fact that birth weight is far and away the most powerful predictor of the chances of infant survival. As crucial as this issue is, the implications of birth weight for child health extend well beyond chances of surviving the first year of life. Low birth weight is associated with a host of childhood health problems and with poorer educational achievement and family functioning. The fact Mexican Americans are able to achieve such remarkably favorable birth outcomes in a disadvantaged social and economic context suggests the operation of ameliorating factors which offset what would otherwise be quite adverse consequences. Identification of these factors and specification of the mechanisms through which they operate would provide extremely valuable information in support of efforts to promote the health and well-being of infants and children of all racial and ethnic groups. Using a recently available nationally representative data set (the National Maternal and Infant Health Survey), testing of hypotheses will proceed first through analysis of birth weight as a continuous variable using conventional regression methods. When attention is shifted to the components of low birth weight multinomial logistic regression models will be applied to the analysis race and ethnic differentials in the risk of a preterm delivery and intrauterine growth retardation.